Abstract
Over-the-air (OTA) advanced driver assistance systems (ADAS) updates may require drivers to adjust their mental models to maintain appropriate system use. Based on the magnitude of change to the ADAS, hands-on experience with the updated system may not require consumer education or training. In this study, 96 drivers experienced adaptive cruise control in the NADS-1 and received an OTA update with either pedestrian detection (PD) or traffic jam (TJ). Drivers that experienced PD had a higher confidence in their understanding of PD whereas drivers that experienced TJ, had a better understanding of TJ. However, their confidence was not correlated with their understanding of the system for either TJ or PD. Although hands-on experience with a new system may improve driver’s understanding or confidence in their understanding of the system, consumer education may strengthen the relationship between their understanding and confidence in their understanding of ADAS.
Keywords
A drivers’ mental model of advanced driver assistance systems (ADAS) and confidence in their mental model shapes how they interact with ADAS while driving (Gaspar et al., 2021). Incorrect mental models may lead to misinterpretation of threats from the surrounding roadway environment, which may diminish the safety benefits of ADAS (Gaspar et al., 2021). Over-the-air (OTA) ADAS updates may require drivers to adjust their mental models to maintain appropriate system use. When system updates occur, drivers might use their existing knowledge and understanding of the previous system to interact with the updated system. Even though OTA updates may be accompanied by description of system changes, they often do not include details about the effect to the driver, such as changes to the driver’s role in performing the driving task. Previous work suggests that drivers do not have a complete understanding of ADAS’s capabilities and limitations (DeGuzman & Donmez, 2021) but that their understanding and use of ADAS can be influenced by training (Zheng et al., 2023). Thus, it is important to understand when and if additional training is needed after an OTA update.
Method
In this study, 96 licensed drivers (MAGE = 42.4 ± 18.0 years; 48 females) completed adaptive cruise control (ACC) training and experienced ACC in the University of Iowa’s National Advanced Driving Simulator (NADS-1). After the first ACC drive, participants were randomly assigned to receive an OTA update with either the pedestrian detection (PD) or traffic jam (TJ) system. Participants completed a subsequent study drive after the OTA update (i.e., “Your vehicle has received the Pedestrian Detection update”). The study drives lasted approximately 20 minute and involved driving through portions of the interstate highway (10 mins) and urban sections (10 mins). Participants encountered low-speed traffic jams, edge-case scenarios (tire in roadway), and pedestrians crossing the street at red lights or while jaywalking.
An initial mental model assessment of ACC was completed after driving with the baseline ACC system (20 true/false items). After the second study drive with the updated system, participants completed a second mental model assessment. The second assessment included questions about both the PD (20 items) and TJ systems (20 items). All mental model assessments also collected self-ratings of confidence on each question. Confidence levels and understanding scores were transformed to range from 0 (low confidence or weak understanding) to 100 (high confidence or strong understanding). A series of one-way ANOVAs were deployed to assess differences in mental model scores or confidence levels between the OTA updated systems. Data are presented as Mean ± Standard Deviation. Multiple comparisons were controlled for using the Benjamini-Hochberg procedure.
Results
A medium strength positive relationship (r =.36, p < 0.001) emerged between drivers’ confidence levels (66.4 ±9.4) and ACC mental model scores (87.2% ± 10.0). Participants that received the PD update reported higher confidence levels (57.4% ± 10.3) compared to the TJ update (45.8% ± 13.2), F(2,93) = 9.438, p < 0.001, η2 = 0.169, but no differences for PD mental model scores, F(2,93) = 1.510, p = 0.226, η2 = 0.031. Participants that received the PD update (53.8% ± 6.6) had a weaker understanding of TJ compared to those in the TJ (60.3% ± 7.6), F(2,93) = 8.022, p < 0.001, η2 = 0.147, but reported similar confidence levels, F(2,93) = 0.076, p = 0.927, η2 = 0.002. No relationships were observed between mental model scores and confidence for either PD or TJ.
Discussion
Drivers that experienced PD reported having higher confidence in their understanding of PD compared to those that received the TJ update. Drivers that experienced TJ had a better understanding of TJ but had similar confidence in their understanding of TJ compared to the those that did not experience the system. Unfortunately, there was no relationship between their confidence and their understanding of PD or TJ. Increased confidence or self-efficacy using ADAS may increase drivers’ intention to use ADAS (Seuwou, 2021).
These observed differences (i.e., relationship between understanding and confidence) between ACC and the OTA updates may have resulted from drivers completing ACC training prior to driving with ACC whereas the OTA updates did not include training or consumer education. The relationship between confidence and mental model speaks to the calibration between actual and perceived understanding. Miscalibration could lead to overconfidence and the misuse or disuse of ADAS. Consumer education or training prior to hands-on experience with PD or TJ, may establish a relationship between drivers’ understanding of the system and confidence in their understanding of the system.
